Medical imaging may be used to analyze internal patient anatomy and pathology. This analysis may facilitate a diagnosis, for example by detecting a tumor, leakage after stent placement, iron deposition in the liver, a fracture, etc. More specifically, an image of a tumor may be analyzed to determine features of the tumor (e.g., a size, shape and/or location), and a diagnosis may be determined based in part on these features.
The determination of internal features based on medical images is referred to as feature extraction. Conventional feature extraction systems are limited in their ability to efficiently detect texture, boundaries and/or densities. Efficient systems to improve feature extraction and, as a result, subsequent diagnoses, are desired.